1. Technical Field
The present disclosure relates to an image sensor and, more particularly, to a lossless non-linear analog gain controller in a CMOS image sensor (CIS) type image sensor.
2. Discussion of Related Art
CMOS image sensors are commonly used in portable cameras, digital still cameras, web cameras and the like to convert an image into a digital signal. The digital image signal that is output from the CMOS image sensor contains red, green, and blue image data. The digital image signal is processed to drive a display device such as a liquid crystal display.
A CMOS image sensor uses a correlated double sampling (CDS) method and uses a ramp signal to generate a digital signal from the difference between a reset signal and an image signal which are sampled according to the CDS method. That is, the CMOS image sensor picks up the difference between the reset signal and the image signal that varies based on illumination conditions, and generates a digital code corresponding to the difference. The digital code varies depending on the slope of the ramp signal under the same illumination conditions. To maintain constant luminosity or brightness under the same illumination conditions, the ramp signal has to be constant when the image acquired by the CMOS image sensor is displayed on a display device.
However, when the scene's dynamic range is equal to the sensor's dynamic range, a method of controlling an analog gain by keeping the ramp signal constant may cause data loss, for example, by saturating highlights when increasing the analog gain to brighten shadows.
Many CMOS image sensors have a pixel array which exhibits a linear response to light. Recently, sensors including non-linear analog pixels or smart pixels have been introduced to increase the dynamic range of the sensors. However, a log sensor suffers from excessive fixed pattern noise (FPN) when transistors in the pixel array are mismatched. Noise sensitivity and image quality may deteriorate since the CDS method cannot be used, due to the continuous operation of the log sensor. In addition, the pixel size becomes large due to the number of transistors in the pixel array.